# Federal Contractors Assess Anthropic AI Risks

MeriTalk articles describe uncertainty around federal use of Anthropic’s Claude following the company’s conflict with the Pentagon over autonomous weapons and mass surveillance. They warn that restricting AI vendors may not eliminate federal employees’ demand for AI tools and could push use of unapproved applications outside agency security visibility. No solicitation, award, funding, contract value, or procurement timeline is identified. For procurement teams and contractors, the issue is how to use AI on government work while accounting for authorization, supply-chain, and shadow-AI risks.

- Contractors using Claude on federal work should assess their current use and potential authorization or supply-chain exposure, and consult applicable agency guidance before relying on the model.
- Agencies and industry may face demand for secure, usable AI tools alongside usage monitoring and governance capabilities; the articles identify this as a potential market need, not a funded procurement opportunity.
- No specific contract or deadline is reported, so the signals do not establish a current solicitation or mandatory new compliance requirement.

**Jurisdictions:** federal
**Industries:** Information Technology, Defense & Military
**Topics:** Artificial Intelligence, Cybersecurity
**Published:** October 29, 2026

### Government Entities
- U.S. Department of Defense (DoD)
- U.S. federal government

### Vendors
- Anthropic (AI model provider)

### Sources
- [Shadow Boxing: Anthropic and Uncle Sam’s AI Problem – MeriTalk](https://www.meritalk.com/shadow-boxing-anthropic-and-uncle-sams-ai-problem) - MeriTalk
- [Anthropic’s IPO: Uncle Sam Wants His AI Both Ways – MeriTalk](https://www.meritalk.com/anthropics-ipo-uncle-sam-wants-his-ai-both-ways) - MeriTalk